 (3).png)
1 hour ago
1
ARTICLE AD BOX
OpenAI is retiring GPT-4o and several older models from ChatGPT today, shifting focus to newer GPT-5 systems with improved personality, creativity and customisation. The change affects ChatGPT only, while API access remains unchanged.
OpenAI is phasing out several of its older artificial intelligence models from ChatGPT today, marking the end of GPT-4o and a number of related systems as the company shifts its focus to newer generations.
In a blog update dated 29 January, the company confirmed that, from 13 February 2026, GPT-4o, GPT-4.1, GPT-4.1 mini and o4-mini will no longer be available inside ChatGPT. Previously announced retirements of GPT-5 Instant and GPT-5 Thinking are also taking effect. OpenAI clarified that these changes apply only to ChatGPT and that its API offerings remain unaffected for now.
Why is OpenAI phasing out the older AI models?
Among the models being discontinued, GPT-4o received particular attention.
OpenAI said it had initially planned to remove GPT-4o earlier, but restored access after feedback from Plus and Pro subscribers. Some users reported that they relied on the model for creative brainstorming and preferred its more natural, conversational tone.
That input, the company explained, influenced the design of its newer GPT-5.1 and GPT-5.2 systems, which aim to replicate and improve on GPT-4o’s warmer personality while offering greater flexibility.
More control over tone and style
The latest models introduce additional customisation features that allow people to adjust how ChatGPT communicates. Users can select base styles, such as a friendlier tone, and fine-tune elements like warmth and enthusiasm.
OpenAI said the goal is to give users more control over how the assistant feels to interact with, not only how it performs tasks.
Usage shifts to newer models
According to the company, adoption data played a major role in the decision. It claims that most ChatGPT activity has already moved to GPT-5.2, with only around 0.1 per cent of daily users still choosing GPT-4o.
With demand concentrated on the newer systems, OpenAI said maintaining older models no longer made sense.
Broader changes ahead
Beyond the retirements, OpenAI outlined further updates aimed at improving ChatGPT’s personality and creativity while reducing what users have described as overly cautious or unnecessary refusals.
The company also said it is working towards an experience tailored for adults, centred on greater freedom and choice within safety limits. As part of that effort, age prediction tools for under-18 users have been introduced in most regions.
A difficult transition for some users
OpenAI acknowledged that removing familiar models may be frustrating for a small group of users who still depend on them. However, it argued that retiring legacy systems allows it to focus resources on advancing the models most people use today.
For many ChatGPT users, the shift will likely go unnoticed. For others who favoured GPT-4o’s distinct style, today marks the end of an era.
Key Takeaways
- OpenAI is discontinuing older models to concentrate resources on newer, more popular versions.
- New models offer enhanced customization features to improve user interaction and experience.
- Adoption data showed minimal usage of older models, justifying their retirement.
English (US) ·